[A combination of horizontal/vertical prism bar].

BACKGROUND

Calibration and positioning of prism bars is not well standardized. When both the horizontal and vertical prism bars are used, they are frequently held back-to-back in front of one eye; the prism bar then reversed from its usual position will give an erroneous measurement.

METHODS

We used an optical bench to investigate calibration and errors from improper positioning of prism bars manufactured by the companies R. O. Gulden and Luneau. We also urged Gulden to redesign its prism bars so that they can be held back-to-back and slid vertically with respect to one another, held in alignment by grooves along the edges of the flat surfaces of the bars. The combination horizontal/vertical prism bar thus created demands that the horizontal prism bar be calibrated precisely in the frontal plane position, and the new vertical prism bar be calibrated in the Prentice position.

RESULTS

Gulden's new combination horizontal/vertical prism bar can be used without significant error (measured deviations within +/- 0.5 delta of labelled values). Gulden's old vertical prism bar has always been calibrated in the Prentice position (a surprise to the company). Luneau's horizontal and vertical prism bars are both calibrated close to the frontal plane position (within +/- 0.4 delta). Improper positioning demonstrated an increasing error with larger prisms. Luneau's 25 delta segment measured in the Prentice position 27.8 delta and the 40 delta segment 67.5 delta; the 25 delta segment of Gulden's old and new vertical prism bars measured 23.2 delta in the frontal plane position.

CONCLUSIONS

Gulden's vertical prism bars, calibrated in the Prentice position, whether used in combination with the horizontal prism bar or by themselves, should always be held with the flat surface toward the examiner. Luneau's horizontal and vertical prism bars should be held one in front of each eye when used simultaneously. As a consequence, neither eye is looking directly at the fixation object, and strictly defining primary and secondary deviations is not possible.